---
title: Wasup Gold Annotator
emoji: ✍️
colorFrom: yellow
colorTo: red
sdk: gradio
sdk_version: 6.0.2
app_file: app.py
pinned: false
---

# Wasup Gold Annotator

Gradio tool for editing temporal subtask segments in `macrodata/whats_going_on_bench`.

The app loads the existing gold subtasks as a starting point. Saving appends one JSON object per episode to an annotations JSONL file. Each saved row contains both the old and revised segments so the benchmark can later evaluate against old gold, new gold, or both.